Re: More entity confusion and my opinion on the right way

[David Carlisle <davidc@xxxxxxxxx>]

> since the & can appear in the literal form, shouldn't it remain in the
> literal form when it is written back out from the result tree?

but it is not in a CDATA section there.

If the XSL stylesheet has <![CDATA[&foo;]]> then that, since this is
 CDATA,  is the 5 characters. & f o o ;.

So how is that to be written to the result tree?
xt currently does
&amp;foo;
which correctly represents those characters.
It could I suppose do <![CDATA[&foo;]]> again, but it would be hard
for it to know good places to start and stop CDATA sections.

It can't write out &foo; as that is not the five characters that
it is trying to output as specified in the original CDATA section,
it is something else, or an error, depending on whatever foo is defined
to be in the DTD for the output tree.

DSSSL (or rather jade) has
(declare-flow-object-class formatting-instruction
  "UNREGISTERED::James Clark//Flow Object Class::formatting-instruction")

to allow markup to be output. I would guess that what the XSL folks want
to avoid is making it easy to output non well formed XML from XSL (which
seems reasonable).

David


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list